15 * speed depends on source allignment
16 * destination allignment is secondary
17 * byte-at-a-time foreward copy to
18 * get source (A1) alligned.
29 * quad-long-at-a-time forward copy
41 * cleanup byte-at-a-time
57 * everything the same, but
65 * byte-at-a-time backward copy to
66 * get source (A1) alligned.
77 * quad-long-at-a-time backward copy
89 * cleanup byte-at-a-time backward